This project seeks to determine the impact of taxes, wages, incomes, social benefits and social legislation on the life cycle dynamics of fertility, labor supply, marital status and wage growth of modern Swedish women. To achieve this objective, we will create a new longitudinal data set by matching data from a richly detailed demographic study (the Swedish Fertility Survey) with tax records on incomes, benefits and assets to be obtained from Statistics Sweden. By combining accurate individual demographic event histories for five (five year) cohorts of Swedish women with individual data on taxes, transfers and incomes, we will create a unique body of data ideally suited to explore the impact of social programs and wages on demographic dynamics. En route to achieving these empirical objectives, we will develop and implement new statistical models for the analysis of dynamic demographic processes. Two frameworks designed to operationalize a broad class of behavioral choice models will be developed. One is for discrete time and the other is for continuous time. Computational procedures for implementing these frameworks will be developed in the form of general purpose algorithms that will be useful for the analysis of related dynamic problems. We also propose to update and expand our (CTM) multistate duration program. We will consider basic theoretical and empirical issues of model validity. Theoretical work on the identifiability of multistate duration models will be conducted. We will continue our work on developing goodness of fit and other criteria to assess the empirical performance of fitted duration models We will continue our work on evaluating alternative model selection criteria for the class of multistate models that will be used in our empirical investigation.
